Trajectories of risk and resilience represent contrasting pathways from early adolescence to young adulthood. Persistent social disadvantage and continuing patterns of maladaptive behavior in adolescence tend to negatively affect health and well-being in the young adult years through impaired education, unstable relationships and work. Even so, adolescents at risk frequently manage to find escape routes to greater opportunity and positive adaptations, whereas other youth lose the advantage of their social origins and become increasingly at risk. Little is known about the contextual, behavioral, and biological mechanisms by which trajectories of risk shape health and well-being into the young adult years, or about the mechanisms by which young people escape from the risk process to trajectories of resilience, especially minorities. Using panel data from Add Health, subproject n proposes a research plan defined by three interrelated objectives that are consistent with the integrative life course theoretical framework of the Program Project: (1) to interrelate risk trajectories[unreadable]based upon sequences of contextual, behavioral, and biological risk factors-the life course, and patterns of health from adolescence into young adulthood, with emphasis on their linking mechanisms;(2) to investigate mechanisms that enhance a trajectory of resilience from adolescent adversity into the 30s, including individual resources, social support and life transitions;and (3) to assess the generality of these life patterns by social contexts, subpopulations, including the genetics subsample. The data will come from Add Health's in-school survey (Wave I in-home interview with Ss &caregiver (1995), Waves n (1996) and HI (2001/2002), and Wave IV, a proposed interview with Ss in 2007-2008. Collaborations are planned with the Harris subproject on family formation trajectories and with the Williams, Hewitt, and Guo subprojects on the assessment of gene-environment effects on risk-resilience and health, and on young adult biomarkers of distress. Modes of data analysis will be contingent on the specific question and available data, but the research plan calls for both variable-based and person-centered strategies, including standard linear regression, and latent growth and mixture models.